Subject:FEL Upgrade Project Weekly Brief - October 30-November 3, 2006

MEMORANDUM

To:Distribution

From:F. Dylla/grn

Subject:FEL Upgrade Project Weekly Brief - October 30-November 3, 2006

Date:November 3, 2006

Highlights:

This is a very exciting weekly report to write for the JLab FEL team. Last week we omitted our usual end of week report in place of our brief announcement distributed last Thursday when we broke the 10 kW milestone at short wavelengths by delivering 11.6 kW of average power at 1.6 microns. In this report we provide the details behind last week’s achievement, in addition to the summarizing this week’s results. We are pleased to report that we pushed the power record to14.26 kW on Monday. This power was obtained with the accelerator driver operating at 8.0 mA which means the lasing efficiency was as high (1.5%) as the value we previously measured at low currents.

The cry-cooled outcoupler mirror which has been responsible for this high power operation has performed admirably. We see no signs of optics limited operation.

We spent some time conditioning the accelerator to run at higher currents (> 8 mA), and found that presently we are limited to sustained operation in the 8-8.5mA range by vacuum excursions in the waveguide inputs to the injector cryounit. High current operation and rf conditioning of the cryounit over the last two weeks has allowed us to raise these limits. We are generating further evidence for the improved power coupler designs we are incorporating in our next version of this important hardware (the injector cryounit-see SRF Report below). While we were exploring the limits to high current operation, the photogun operation was interrupted by a gun arc and we restored the gun to normal operation by a heat clean cycle. Prior to this interruption, we obtained a new record of charge extraction (525 Coulombs) from the photocathode between cesiation cycles.

Details of the operations during the last two weeks and the performance of the accelerator, rf, photogun and optics systems are given in the report below. The plot below shows extended operation of the machine above 10 kW on Tuesday just prior to obtaining the 14 kW power record.

Additional plots are shown in the I&C report below.

Today we are exploring additional extended operation of the machine at the 10 kW level and testing the performance of the Optical Transport System (OTS) to the user labs. Next week, FEL operations will be devoted to THz radiation production and an attempt to test a high sensitivity THz camera on loan from the US Army Night Vision Lab.

Management:

FEL Management attention for the last two weeks was totally focused on

the pleasant task of coaching the FEL team past the 10 kW level. The team needs little coaching but the hardware always seems to appreciate another set of eyes.

A number of our colleagues in the business spent the week at the DEPS meeting in Albuquerque. George Neil represented the FEL team and Roy Whitney presented a paper

on our conceptual studies of airframe platforms for an FEL. We are pleased to note that DEPS presented George with a Fellow Award to recognize his career long contributions to directed energy. George graciously accepted award on behalf of all of his colleagues who have helped him design, test and operate the JLab FELs, and to thank the conference organizers he aced the Raytheon Poker Challenge for the two night competition.

Coordination meetings were held this week with Axion study group from Hampton/Yale

and the AMO study group from ODU.

Congratulations to Jianxun Yan for winning the Isamu Abe prize at last weeks PCaPAC conference and to Chris Tennant who passed his Ph.D defense at the College of William Mary on Oct. 26th. Chris will be joining the FEL team as a newly minted post-doc on Nov. 16th.

Operations:

The past two weeks have been consumed with optimizing and measuring the performance of the laser with the cryomirror. Recall that during the week of Oct. 16-20 we obtained good efficiency early in the week at low current and up to a record 7.5 kW with 6 mA of current but could not manage to optimize the laser at high power. Later in the week the efficiency fell somewhat.
The drive to high power consisted of three goals. The first was to increase the efficiency at low current. The second was to preserve this efficiency at high current. The third was to increase the current as high as possible.
The week of Oct. 23-27 started with an effort to reduce shifts in the wiggler vacuum chamber. We first optimized the FEL and got up to 7.8 kW. We also did a wavelength scan to figure out what gap was optimal for lasing. We then locked the wiggler at this gap and installed cooling plates on the wiggler chamber to keep it from moving while running high current. We tested this out on Tuesday and found that the chamber movement was reduced by more than one order of magnitude. The center of the wiggler chamber now only moved by 130 microns when running over 6 mA for extended periods. The position would recover more quickly as well and the vacuum rise was not as bad as before. On Wednesday we were able to recover the efficiency we had demonstrated the previous week. We thus met our first goal in the drive to high power. When we ramped up the current the efficiency dropped. We were still able to obtain

9.4 kW at 6.7 mA. This was the first time we had exceeded the CW power record obtained at 5.75 microns in 2004. We spent the evening doing coupler conditioning in the injector so that we could push the current up on Thursday.
After tuning up the machine Thursday morning we obtained a good lasing configura-tion with close to 1.8 kW/mA efficiency at low current. We had good success optimizing the machine at high current once we backed off the HR mirror heater, increasing the Rayleigh range. Though the laser efficiency at low current was higher with the higher HR heater setting, it was not possible to optimize the mirrors because intercepted power would steer the mirrors too fast to allow optimization. Once the mirrors were fairly stable we were able to optimize the system fully and obtain 11.7 kW at 7 mA. This is almost as high an efficiency as at low current. We therefore had met our second goal.
To get to higher power we needed higher current so we worked on Friday to get higher current from the accelerator. We were able to gradually push the current up to 8.8 mA and could operate for extended periods at 8.5 mA. Note that this is the highest current we had ever run with the zone 3 cryomodule installed.
Monday, having achieved all three of our goals, we were ready to push for high power. We ran for extended periods at 8 mA and gradually optimized the system until we obtained 14.26 kW at 8 mA. At low current we obtained 2.2 kW at 1.25 mA. Within the uncertainty in our power and current measurements this is the same efficiency of 1.55%. It is actually slightly higher at the higher current so there is no evidence for any degradation in efficiency at a current of 8 mA. The efficiency at 3.75 mA was also 1.5%. At this current the laser is quite stable. We ran at 5.5 kW for 20 minutes with no adjust-ments at all. At very high power the mirrors tend to drift so that it is hard to hold the power for long periods of time. Vacuum levels also rise and cause vacuum trips after about an hour of running at the highest current. This is slowly cleaning up after running enough ampere hours.
Clearly the only way to increase the power from 14.26 kW was to further increase the current so we worked on Tuesday to try for higher current. Though we could reliably run at 8 mA and could run for extended periods at 8.5 mA we could not push the current to 9 mA for more than a few seconds. We decided to try to run for a while at high power but an arc occurred in the gun, shutting down operations for the day. We made a new cathode, HV processed it, and were ready to run by Thursday morning. While we were down on Wednesday we realigned the optical transport so that we could send the power upstairs. The transport line was pumping down on Thursday so we could not lase at high power. We decide to run moderate current (6 mA) for several hours to test a hypothesis that running a lot of current right after making the cathode increases the cathode lifetime.
We believe higher power is achievable if we can increase the current. It is also achievable if we use an output coupler with smaller output coupling. We may be able to get 18 kW at 8 mA with the 11% output coupler. At some point there will be some optical limit but there is no sign of it at 14 kW. Conditioning may raise the attainable current to 10 mA but we may have to change the cryounit window design to get more than 8.5 mA. Operation at higher gun voltage might also be an advantage.

WBS 4 (Injector):

The photocathode gun delivered a record 525 Coulombs and 25 hours of CW beam time with an average current of 7 mA without a re-cesiation. The last record was 450 Coulombs and 44 hours of CW beam time. The photocathode delivers routinely an average of 25 Coulombs per hour and has demonstrated 8 mA CW stable operation for periods as long as 30 minutes. Longer periods are limited by quarter cryounit waveguide vacuum trips and steady vacuum rise downstream of the cryounit. Higher currents are being limited by vacuum events that damage the photocathode with a visible divot and total loss of QE. However, the photocathode has been recovered twice from these events in the last two weeks. In fact, the photocathode lifetime was cut short at the 525 Coulombs level due to one of these events, otherwise the lifetime should have been much more longer since the Drive Laser power request barely increased from 27% to 30% in several days of running 7 mA+ and extracting over 100 Coulombs per day of operations. On Wednesday we activated the GaAs wafer into a photocathode to recover from the last QE loss event. The ops team was able to run steadily at 6 mA for over two hours almost immediately after recovering the photocathode QE and extracted the usual 25 Coulombs per hour. We are brainstorming to install diagnostic equipment in the injector to look for the cause of these events and eventually learn to overcome them.
We also received the stand to mount the new gun chamber and ordered the DC motor and gears for the new stalk retraction mechanism. Don Bullard and Nimel Theodore mounted a new electrode in the high voltage test stand and baked this system. Nimel started HV conditioning this electrode on Thursday which he previously coated with the field emission suppression film. Don has also made progress polishing the anode plate for the new gun chamber completing 300 grit. Deepesh Kumar, our ODU student, has made significant progress understanding and modeling the FEL injector in PARMELA and running the optimization routine.

Gun HVPS - It is fully operational. It reliably delivered over 8 ma of current whenever requested without faulting or being noisy. RF cavity waveguide trips were the primary reason for not operating at even higher currents.

WBS 5 SRF:

High-current cryomodule update:
Another small step forward was achieved this week with thetesting of our tooling for producing high average power capable RF windows. Tom Elliot and Mircea Stirbet of the SRF institute ran thetooling through a full furnace cycle and we took the opportunity totest a new "brazeless" assembly technique that uses temperature andpressure to diffusion bond the ceramic directly to the cooling ring (see photo). The result was successful first time out, producing anice looking leak-tight seal. This is also the first demonstration ofthe rugged pre-stressed ceramic window design at JLab. The method haspreviously been used at SLAC and LBNL for MW class CW windows at 476MHz and 700 MHz. More tests including more conventional brazed butstill pre-stressed assemblies are planned.
The 1.5 GHz cavity prototypes are finally ready to move forwardagain now that the batch of highest-priority 12 GeV related cavitywork has cleared the e-beam welder. The 750 MHz cavity dies have beentested using aluminum and copper sheets and once the frequencies havebeen checked we can also proceed with forming the first 750 MHz niobium cells. Work has started on designing the dies for the 750 MHzcavity waveguide end groups.
We have a handful of different candidate HOM absorber materialsnow characterized, several of which may also be usable at cryogenictemperatures. Work is proceeding to optimize the load geometryforhigh power and tolerance against variations in material propertiesfrom batch to batch.
We have tentatively scheduled our 750 MHz cryomodule conceptualdesign review for the third week in November, date to be confirmedsubject to reviewer availability.
Unfortunately despite this good progress we had a set-back thisweek when the majority of our engineering resources was pulled towork on the 12 GeV project. We think we can maintain the ongoingprototype activities using primarily institute staff but this will impact plans for the 750 MHz module detailed design and possible Admiral replacement and new cryounit activities.

Progress on the FEL/AES 100 mA Injector Cryounit
-Warm-to-Cold beampipes: Components have been received; final assembly will begin later this month when resources are expected to become available.

-HOM feed-through: Prototype parts are staged for final braze assembly; testing with fundamental cavity will follow.

-FPC: Visual and dimensional check with CMM has been completed, some additional machining of parts was necessary for fit up, low level rf testing will follow final component assembly.

-Cavity string: Assembly tooling drawings are signed-off and out for fabrication bids. --Resource availability continues to pace schedule.

WBS 6 (RF):

RF -The RF system ran quite well these past weeks. It reliably provided power to injector and linac cavities to accelerate and control over 8 ma of e-beam current. The only limitation was cavity waveguide vacuum trips in the injector at about 8.5 ma, but this is improving with RF processing and as higher and higher beam current is run. The cavity waveguides are cleaning up slowly as there is a small helium leak into the beam line from the cryo-cooled out-coupler mirror. All in all, it was a very good week as the FEL lased as high as 14.2 kW and easily ran above 10 kW for minutes on end.

WBS 8 (Instrumentation):

These last 2 weeks have been both exciting and exhausting. The entire group has been working hard to both support operations in our high power push and to make sure we are ready for the December shutdown. In addition to record laser power and currents we were able to complete the installation and certification of the new Laser Personnel Safety System in Lab 1. Coordination activities have also begun for the upcoming scheduled downtime in December. The EES group has been contacted for scheduling preventive maintenance on the trim racks along with testing all the interlocks on the shunt module crates, box supplies and flow switches. Another acid wash, depending on the flow rates, could be in order for the box supplies. Also, EES will provide some PM on our VME and CAMAC crates and modules during the down. We have moved installation efforts into User Lab 6 with prepping the LPSS cabinet with current modifications (rewiring and adding connectors, connecting grounds and labeling). The mirror cassette bypass box will be modified to accommodate the added length to the actual position of the mirror cassette can.
Last week the online FEL performance graph had to be modified to display power levels over 10kW since the old graph was limited to 10kW full scale. This was a welcomed issue when the FEL power disappeared off scale. An extension to the graph was included to allow for our power records of 11.7kW and then 14.3kW (see figure above) to be displayed as well as any future records beyond that. Also, the epics-to-devlore data archiving process has been running in the background now for a few months. With the excitement of high power lasing and high current running we decided to calculate the total charge from the gun and the total energy produced at 1.6 micron for the entire month of October. (See figure above). This represents the first of the monthly system performance analysis. Also, a process was written to generate a plot of the vacuum readbacks for the whole vacuum system. Included in this plot is, not only the pressure readings, but also a mini-8-hour trend for each pump (see figure above). The vacuum plot updates each minute and is viewable online.
In support of the LIPSS experiment, FFT calculations are being performed on the data from their camera. There is a noise signal on the camera image that is much greater than its spec. claims. Hopefully, understanding the noise spectrum will help the manufacturer diagnose the camera.
After working carefully with the PCB manufacturer to make sure the design was correct, the Single Board IOC revision was ordered. Numerous manufacturing issues were resolved to better understand the process in which the PCBs are manufactured. This will enable us to better design boards before they go out for quotes and in turn alleviate unnecessary delays and complications. We also ordered the backplane for the Beam Viewer Crate upgrade. Both of these boards are expected to arrive by the end of next week. Many of the components to assemble these boards and others have arrived and a few remaining parts still remain to be ordered. The quotes were requested for the front panels of the RTD board and the Beam Viewer Control Card. We are ordering production level quantities of these two boards since the front panel layout will not change.
We've spent some time working on the next revision of the BPM electronics. Wehave currently tested a setup to provide a self trigger andalsowe are looking at the sampling frequency we will be able to achieve with the SBIOC. Some work has been doneto layout the PoE (Power over Ethernet) frontend that we would like to implement in the nextrevision. Thepowered device controller and theDC/DC converter have been specified and the configuration for these two is being examined. We're also working with the appropriate groups to have the new Sextapole Power Supplies ready for installation during the December down. The required parts are being acquired and accounted for so we can be ready to go right away.
Work continues on the Single board IOC. The FPGA was connected with ADC AD7655 and the code was programmed for the AD7655. The AD7655 is soldered on the general purpose board and wired in the parallel mode. The communication tests between the FPGA and AD7655 will continue. Congratulations to Jianxun Yan for winning the Isamu Abe prize at last weeks PCaPAC conference for this work!
Some work was done on wiggler controls and diagnostics. A new test version for the Wiggler software was established. This is being tested as much as possible on a development IOC before use with the actual Wiggler controller.